Bug 16451 - hugin crashes at launch after mga5 update
Summary: hugin crashes at launch after mga5 update
Status: RESOLVED FIXED
Alias: None
Product: Mageia
Classification: Unclassified
Component: RPM Packages (show other bugs)
Version: 5
Hardware: x86_64 Linux
Priority: Normal critical
Target Milestone: ---
Assignee: Mageia Bug Squad
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on: 16560
Blocks:
  Show dependency treegraph
 
Reported: 2015-07-23 10:27 CEST by Bruno Cornec
Modified: 2016-03-26 17:20 CET (History)
3 users (show)

See Also:
Source RPM: hugin-2014.0.0-0.rc4.5.mga5.src.rpm
CVE:
Status comment:


Attachments
.hugin file causing the crash (4.87 KB, text/plain)
2016-03-26 17:17 CET, Bruno Cornec
Details

Description Bruno Cornec 2015-07-23 10:27:29 CEST
Description of problem:

hugin crahses at launch after mg5 update

Version-Release number of selected component (if applicable):
hugin-2014.0.0-0.rc4.5.mga5

How reproducible:

Always

Steps to Reproduce:
1. Launch hugin at cli
2. Get the following messages:

(hugin:12029): Gtk-CRITICAL **: IA__gtk_widget_set_size_request: assertion 'height >= -1' failed

(hugin:12029): GdkPixbuf-CRITICAL **: gdk_pixbuf_new: assertion 'width > 0' failed

(hugin:12029): Gtk-CRITICAL **: IA__gtk_widget_set_size_request: assertion 'height >= -1' failed

(hugin:12029): GdkPixbuf-CRITICAL **: gdk_pixbuf_new: assertion 'width > 0' failed

(hugin:12029): Gtk-CRITICAL **: IA__gtk_widget_set_size_request: assertion 'height >= -1' failed

(hugin:12029): Gtk-CRITICAL **: IA__gtk_widget_set_size_request: assertion 'height >= -1' failed

(hugin:12029): Gtk-CRITICAL **: IA__gtk_widget_set_size_request: assertion 'height >= -1' failed

(hugin:12029): Gtk-CRITICAL **: IA__gtk_widget_set_size_request: assertion 'height >= -1' failed

(hugin:12029): Gtk-CRITICAL **: IA__gtk_widget_set_size_request: assertion 'height >= -1' failed

(hugin:12029): Gtk-CRITICAL **: IA__gtk_widget_set_size_request: assertion 'height >= -1' failed

(hugin:12029): Gtk-CRITICAL **: IA__gtk_widget_set_size_request: assertion 'height >= -1' failed

(hugin:12029): Gtk-CRITICAL **: IA__gtk_widget_set_size_request: assertion 'height >= -1' failed

(hugin:12029): Gtk-CRITICAL **: IA__gtk_widget_set_size_request: assertion 'height >= -1' failed

(hugin:12029): Gtk-CRITICAL **: IA__gtk_widget_set_size_request: assertion 'height >= -1' failed
/usr/share/hugin/data/plugins/top_five.py
   CAT:Control Points
   NAM:keep 5 CPs per image pair
/usr/share/hugin/data/plugins/shooting_pattern.py
   CAT:initial distribution
   NAM:6-1-1 Shooting Pattern
/usr/share/hugin/data/plugins/crop_cp.py
   CAT:Control Points
   NAM:Crop Control Points
/usr/share/hugin/data/plugins/woa.py
   CAT:Control Points
   NAM:Warped Overlap Analysis
Segmentation fault

Reproducible: 

Steps to Reproduce:
Comment 1 Rémi Verschelde 2015-07-23 12:03:10 CEST
CC'ing the most active hugin committers.

@Bruno: You say it crashes after update. Did you mean upgrade (i.e. the problem is there since Mageia 5 was released), or was it really introduced by a recently updated package? If so, could you provide the output of rpm -qa --last | head -n 100?

CC: (none) => balcaen.john, fundawang
Summary: hugin crahses at launch after mg5 update => hugin crashes at launch after mga5 update

Comment 2 Rémi Verschelde 2015-07-23 12:04:52 CEST
While fixing this bug, hugin should also be upgraded to the final 2014.0.0 release instead of RC4.
Comment 3 Rémi Verschelde 2015-07-23 12:08:32 CEST
Tested locally on Mageia 5 64bit, and hugin starts fine. I get the exact same console output as in comment 0, apart from the Segmentation fault.
Comment 4 Rémi Verschelde 2015-07-23 12:40:08 CEST
Could you try to run hugin through gdb to get a stacktrace? See also https://wiki.mageia.org/en/Debugging_software_crashes
Comment 5 Bruno Cornec 2015-07-27 05:42:23 CEST
(In reply to Rémi Verschelde from comment #1)
> @Bruno: You say it crashes after update. Did you mean upgrade (i.e. the
> problem is there since Mageia 5 was released), or was it really introduced
> by a recently updated package? If so, could you provide the output of rpm
> -qa --last | head -n 100?

I hadn't tested hugin before. So I upgraded from mga4 to mga5, then applied all updates, and needed to run hugin so launched it and realized it was crashing. However, on my laptop, it doesn't crash (also updated from mga4 to mga5). 

I've had issue during my upgrade, so the rpm --last is not really relevant. Anyway here is the hugin line:
hugin-2014.0.0-0.rc4.5.mga5.x86_64            mar. 07 juil. 2015 01:46:21 CEST

As I'm away from home for 3 weeks, I tried to relaunch it with ssh, and it's slow.
I seem to have also issues to add the debginfo pkg, but I get the following with gdb:
Program received signal SIGSEGV, Segmentation fault.
0x000000000059be7a in GLPreviewFrame::SetGuiLevel(GuiLevel) ()

(gdb) bt
#0  0x000000000059be7a in GLPreviewFrame::SetGuiLevel(GuiLevel) ()
#1  0x00000000004dc7d5 in MainFrame::SetGuiLevel(GuiLevel) ()
#2  0x00000000004e1103 in MainFrame::MainFrame(wxWindow*, PT::Panorama&) ()
#3  0x00000000004c41b1 in huginApp::OnInit() ()
#4  0x00007ffff560da32 in wxEntry(int&, wchar_t**) () at /lib64/libwx_baseu-3.0.so.0
#5  0x00000000004b11ae in main ()

One difference with mga4 is thatI as using the nvidia proprietary driver, and now I'm using nouveau, so that may be a difference.

I'll be able to give more feedack when I'm back home I think.
Yann Cantin 2015-08-15 14:13:40 CEST

Depends on: (none) => 16560

Comment 6 Yann Cantin 2015-08-20 00:41:02 CEST
There is the 2014 final in updates-testing : hugin-2014.0.0-1.mga5
Not designed to fix this bug, but may be worth a try.

CC: (none) => yann.cantin

Comment 7 Rémi Verschelde 2015-08-30 13:36:31 CEST
Bruno, can you test hugin-2014.0.0-1.mga5 so that we know if it fixes this issue?
Comment 8 Bruno Cornec 2015-10-26 11:13:54 CET
This version continues to crash in a similar manner
Comment 9 Bruno Cornec 2016-03-26 17:17:35 CET
Created attachment 7597 [details]
.hugin file causing the crash
Comment 10 Bruno Cornec 2016-03-26 17:20:19 CET
When renaming the .hugin file in my home directory, a new one is created and the problem disappears.

Attached is the version I had previously which was causing the crash for the record

Status: NEW => RESOLVED
Resolution: (none) => FIXED


Note You need to log in before you can comment on or make changes to this bug.